Effects of succinylacetone on dimethylsulfoxide-mediated induction of heme pathway enzymes in mouse Friend virus-transformed erythroleukemia cells

Experimental Cell Research 154 (1984) 474-484

Effects of Succinylacetone on Dimethylsulfoxide-Mediated Induction of Heme Pathway Enzymes in Mouse Friend Virus-Transformed Erythroleukemia Cells CAROLE BEAUMONT, JEAN-CHARLES DEYBACH, BERNARD GRANDCHAMP, VASCO DA SILVA, HUBERT DE VERNEUIL and YVES NORDMANN* University

Paris 7, Faculty of Medicine Hospital Louis Mourier,

X. Bichat, Department of Biochemistry, 92701 Colombes, France

Heme has been reported to exert a control over its own biosynthesis and to affect the erythroid differentiation process at different sites. In this study, succinylacetone, a powerful inhibitor of d-aminolevulinic acid dehydrase was used to block heme synthesis and to study the effects of heme depletion on the dimethylsulfoxide (DMSO)-mediated induction of the heme pathway enzymes in Friend virus-transformed erythroleukemia cells. The presence of succinylacetone in the medium during the DMSO treatment (1) potentiates the induction of b-aminolevulinic acid synthetase (the first enzyme of the pathway) and this effect is reversed by the addition of exogenous hemin; (2) does not affect the induction of Aminolevulinic acid dehydrase (the second enzyme): (3) prevents the induction of porphobilinogen deaminase (the third enzyme), since no increase could be detected in either the enzyme activity or the immunoreactive protein and this effect could not be reversed by the addition of exogenous hemin; (4) does not affect the induction of ferrochelatase. The possible role of heme or of intermediate metabolites of the pathway on the induction of these enzymes during the erythroid differentiation process is discussed.

All types of mammalian nucleated cells contain hemoproteins and can synthesize heme. There are now strong evidences that the synthesis of heme and apohemoproteins are closely coordinated and are controlled by processes that appear to be tissue specific [l]. It is well documented that, in liver cells, heme plays a central role in the regulation of its own synthesis by controlling the activity of the first and rate-limiting enzyme of the pathway, d-aminolevulinic acid (ALA) synthetase [2]. In erythroid cells, results are less clear-cut, since different effects of heme have been reported, whether considering differentiating cells where heme synthesis undergoes induction or considering already hemoglobinized cells. In the latter situation, namely in reticulocytes, heme controls a translational repressor of protein synthesis [3], it prevents uptake of transferrinbound iron into the cells [4] and it inhibits heme synthesis mainly by direct inhibition of ALA synthetase [5, 63. Heme also affects the erythroid differentiation process, since it has been reported to accelerate hemoglobinization of erythroblasts in bone marrow cell cultures [7] and to enhance the in vitro growth of primitive erythroid progenitor cells [8,9]. However, biochemical studies of the normal erythroid differentiation process are difficult to perform, since bone Copyright @ 1984 by Academic Press, Inc. All rights of rcpraduction in any form reserved 0014-4827/84 $03.00

Succinylacetone

and heme synthesis in Friend cells

475

marrow cells represent a heterogeneous population of cells at different stages of maturation. Therefore, the effect of heme on the control of erythroid differentiation has mostly been studied in Friend erythroleukemia cells which are virustransformed murine erythroid precursor cells. Treatment of these cells with a variety of chemicals, especially dimethylsulfoxide (DMSO), results in synchronous cellular [l&12] and nuclear [13, 141changes which parallel the changes that occur during normal erythropoiesis. Hemin treatment of these cells brings about an increase in globin mRNA levels without committing the cells to terminal differentiation [15, 161; it also increases the activities of ALA dehydrase and porphobilinogen (PBG) deaminase, the second and third enzyme of the heme biosynthetic pathway [17]. Some indirect evidence of a heme-mediated incraese in ALA synthetase in Friend cells also arose from increased incorporation of [‘4C]glycine into heme. But in these cells, no direct measurements of ALA synthetase activity have been performed in conditions affecting heme balance. It has been reported that succinylacetone (SA, 4,6-dioxoheptanoic acid), a potent inhibitor of ALA dehydrase, caused a marked decrease in heme content of Friend cells when added to the culture medium [18]. This drug has also recently been used in cultured chick embryo hepatocytes to study the effect of heme depletion on the induction of ALA synthetase [ 19-2 I]. In this report, succinylacetone was used to block heme synthesis in Friend erythroleukemia cells and to assess the effects of heme depletion on the DMSOmediated induction of heme biosynthetic pathway enzymes. The reversion of these effects by addition of exogenous hemin was also investigated. During the course of this study, a striking effect of SA was observed on two enzymes of heme biosynthetic pathway: (i) it potentiated the DMSO-mediated induction of ALA synthetase and this effect could be reversed by addition of exogenous hemin; (ii) it prevented the DMSO-mediated induction of PBG deaminase and this effect was not related to heme depletion. This last result raises the possibility of a control of PBG deaminase expression by the substrate of this enzyme (porphobilinogen). METHODS Cultures Friend virus-infected cells, clone 745, were grown in suspension in MacCoy’s 5A medium supplemented with 10% fetal calf serum (FCS) and 2 mM glutamine. The cultures were maintained in 5 % COT95% air at 37°C in a humidified incubator. Subcultures were performed every 3-4 days by diluting the suspension to lo5 cells/ml.

Induction

Experiments

Experiments were begun as described by Granick & Sassa [171. Cells for induction were diluted to 10scells/ml to maintain a logarithmic growth rate. 12-16 h later, inducers (DMSO, hemin or both) and SA were added in an equal volume of fresh medium at twice the final desired concentration. DMSO was added to 1.5 % (v/v). A 10e3 M stock solution of hemin was prepared by dissolving hemin in 0.2 N KOH and 100% ethanol (1 : 1, v/v). Hemin was added to the medium at a final concentration of lO-5 M. Exp Cell Res 154 (1984)

476 Beaumont et al. Preparation of Cell Extracts Cells were harvested by centrifugation at 800 g for 5 min and washed twice with phosphate-buffered saline (PBS). The final pellet was resuspended in 50 mM Tris-HCI buffer, pH 7.3 and sonicated twice for 10 sec. An aliquot of the mixture was used for ferrochelatase assay. The remaining lysate was centrifuged at 8 000 g for 5 min and the supematant was used for further analysis. The protein concentration of the cellular extracts was measured before and after centrifugation by the Coomassie-blue binding method (Bio Rad Laboratories, Richmond, Va).

Enzyme Assays ALA synthetase activity was measured by a radiochemical method as described previously [22], using between 0.5 and 1 mg of cellular protein in the incubation mixture. Enzyme activity was expressed as pmoles ALA formed per 30 min per mg protein. ALA dehydrase and PBG deaminase assays were carried out by micromethods [l 11. Ferrochelatase activity was measured by a radiochemical method as reported by Deybach et al. [23], using between 0.05 and 0.1 mg of protein. Enzyme activity was expressed as nmoles mesoheme per hour per mg protein.

Hemoglobin Determination Hemoglobin determinations were performed on freshly prepared cell extracts using benzidine as reported by Conscience et al. [24].

Preparation of SA-Pyrrole SA-pyrrole was prepared from SA by a Knorr-type condensation with ALA according to the method described by Ebert et al. [18] and modified by Brumm & Friedmann [25]. A mixture of 1 ml of 1 mM SA, 1 ml of 10 mM ALA (pH 7) and 1 ml of 1 M sodium acetate buffer (pH 4.6) was boiled for 20 min. The pyrrole was purified by acidifying the boiled mixture to pH 1 with HCl, followed by six extractions with 2 vol of diethyl ether. The ether was removed by heating in a water bath under a stream of nitrogen. The SA-pyrrole was redissolved in 100 mM Tris-HCl buffer, pH 8.2 and the concentration determined spectrophotometrically by the modified Ehrlich reaction using a molar extinction coefficient of 53x lo3 at 555 nm. The visible spectrum of Ehrlich positive material was similar to that of PBG.

Effect of SA and SA-Pyrrole on PBG Deaminase Activity A normal red blood cell-hemolysate was used as enzymatic source of PBG deaminase. Equal volumes of hemolysate (1 : 5 in 100 mM Tris-HCl buffer, pH 8) and of serial dilutions of SA or SApyrrole in the same buffer were first incubated at room temperature for 10 min. The mixtures were then assayed for PBG deaminase activity.

Purification of ALA Dehydrase and PBG Deaminase and Production of Antibodies A DEAE-cellulose chromatography was used as a first step to separate ALA dehydrase from PBG deaminase, starting from the same batch of human erythrocytes. Both enzymes were further purified separately according to the procedures described by Anderson & Desnick [26, 271. Rabbits were immunized against either preparation by repeated subcutaneous injections. Immunoglobulin fraction of rabbit anti-human ALA-dehydrase or of rabbit anti-human PBG deaminase were obtained by a combination of ammonium sulfate precipitation at 45 % saturation and DEAE-cellulose chromato-

awhy WI.

Enzyme Immunoassay (EZA) of ALA Dehydrase The EIA developed for mouse ALA dehydrase utilized the ability of anti-human ALA dehydrase antibodies to cross-react with mouse enzyme. The entire immunoglobulin fraction was labelled with peroxidase by the two-step procedure of Avrameas & Temynck [29]. Specific antibodies were Exp Cell Res IS4 (1984)

Succinylacetone and heme synthesis in Friend cells

477

Table 1. The effects of succinylacetone on hemoglobin content of DMSO-treated Friend cells with or without exogenous hemin Treatment None DMSO 1.5% Hemin lo-’ M DMSO 1.5 %+Hemin lo-’ M DMSO 1.5 %+SA 0.5 mM Without hemin (A) +Hemin 10m5M (B) +Hemin 10m5M (C)

Hemoglobin (ug/mg protein)

Ratio to control

4 101 20 12.5

1 25 5 30

4 38 50

1 10 12.5

Cells were incubated for 5 days with inducers (DMSO, hemin or both) or with DMSO in the presence of succinylacetone (A) without hemin; (B) with hemin for 5 days; or (C) with addition of hemin at day 3 and harvesting on day 5.

isolated by aftinity chromatography on ultrogel-bound human ALA dehydrase and were used at 5 t&ml to coat the bottom of wells in a microtitre plate. Cell extracts (200 ul) were allowed to react with adsorbed antibodies during 3 h at 37°C. After three washings, the peroxidase-labelled antibodies were added and incubated at 4°C overnight. The plate was then washed three times and the enzyme remaining in the wells was determined using 2,2’-azinodi (3-ethyl-benzthiazoline sulfonic acid-6) ammonium salt as a substrate. The absorbance was read at 415 nm. When cell lysates were assayed in these conditions, the results were expressed in arbitrary units (ODlmg protein).

Immunotitration

of PBG Deaminase

Antibodies against human PBG deaminase were shown to cross-react with the mouse enzyme. Therefore they were used for the immunotitration of PBG deaminase present in DMSO-treated cells cultured in the absence or presence of SA. Lysates were adjusted to a same enzyme activity by diluting in PBS containing 5 mg/ml of bovine serum albumin. 40 pl of these Friend-cell lysates were mixed with 40 ul of serial dilutions of anti-PBG deaminase immunoglobulins and left overnight at 4°C. 20 ul of a 10% protein A-Sepharose suspension was added to ensure complete precipitation of immune complexes. After 1 h of incubation, the suspension was centrifuged at 8 000 g for 1 min. The enzyme activity remaining in the supematant was assayed as mentioned above.

RESULTS Effect of SA on Hemoglobin Content and Reversion by Hemin Friend cells, clone 745, treated with 0.5 mM SA for 5 days, exhibited complete inhibition of ALA dehydrase activity as shown by the in vitro assay of the enzyme in cell lysates. Therefore, this drug was used to block heme synthesis and to study the effect of heme depletion on the induction of enzymes of heme biosynthetic pathway and on hemoglobinization of Friend cells treated with 1.5 % DMSO. The cellular accumulation of hemoglobin after 5 days of culture under the various conditions is presented in table 1. Hemin treatment alone resulted in a fivefold increase in hemoglobin content over control cultures, whereas a 25-fold Exp CellRes 154(1984)

478 Beaumont et al. 1 / = ONSO + SA / (NITHOIJT HENIN

i 7

‘t

ONSO P+ HENIN

DN

HEMN

Fig. 1. Effect of succinylacetone on ALA synthetase activity in DMSO-treated Friend cells and reversibility of the effect by hemin. DMSO, hemin, SA (0.5 mM) or combinations of these agents were added at time 0. In one case, hemin (x) was added after 72 h of culture in the presence of DMSO and SA and ALA synthetase activity was determined after another day of growth. Fig. 2. Changes in PBG deaminase activity in Friend cells. Cells were grown in medium containing DMSO, hemin or both, in the absence or in the presence of succinylacetone (0.5 mM).

increase was observed with 1.5 % DMSO. A combination of both inducers resulted in a 30-fold increase. The presence of SA in the medium during the induction experiment with DMSO completely abolished the accumulation of hemoglobin. Exogenous hemin only partially overcame the effect of SA, since a lo-fold increase in hemoglobin content was observed after 5 days with DMSO, hemin and SA as compared with a 30-fold increase when SA was omitted.

Effects of SA on the Induction of Heme Pathway Enzymes In the presence of DMSO alone, ALA synthetase activity increased over a 4day period and then tended to level off. During the first 48 h of the induction experiment in the presence of SA, ALA synthetase activity followed the same time course of induction as with the inducer alone but afterwards a sharp rise occurred in the enzyme activity, leading at 96 h to a sixfold increase over the level obtained with DMSO alone (fig. 1). In contrast, the DMSO-mediated increase of PBG deaminase activity appeared to be completely abolished by the presence of SA in the medium during the induction experiment. The enzyme activity remained constant or even declined to below the activity observed in non-induced cells (fig. 2). Exp CellRes 154(19&t)

Succinylacetone and heme synthesis in Friend cells 479

t 3 + SA

,-

Fig. 3. Changes in ferrochelatase activity in Friend cells. Cells were incubated in medium containing DMSO in the absence or in the presence of succinylacetone (0.5 mM). Fig. 4. Immunotitration of Friend cell PBG deaminase by a rabbit antiserum. Lysates from DMSOtreated Friend cells cultured in the O-O, absence or Ca, presence of 0.5 mSA were adjusted to the same PBG deaminase activity and immunotitration was performed as described under Methods. Each point represents the mean of duplicate measurements.

Finally, ferrochelatase seemed to be induced normally in the presence of SA, since the time course of the DMSO effect on ferrochelatase activity was very similar with or without SA (fig. 3). Effect of SA on Zmmunoreactive ALA Dehydrase The three enzymes studied so far exhibited different responses to the presence of SA in the culture medium throughout the induction experiment. It was therefore of interest to study the effect of SA on the synthesis of ALA dehydrase. Anti-human ALA dehydrase antibodies being available, we developed an enzyme immunoassay (EIA) to quantitate the catalytically-inactive protein accumulated in cytosol. Under the conditions described in Methods, the cross-reactivity of the anti-human ALA dehydrase antibodies appeared sufficient to detect the mouse enzyme. A linear relationship was obtained between the optical density and the cellular protein when the same cellular extract was assayed at different dilutions (not shown). Results obtained by EIA with lysates from DMSO or DMSO- and Exp Cell Res 154 (1984)

480 Beaumont et al. Table 2. ALA dehydrase: levels of enzymatic activity and of immunoreactive protein in DMSO- and DMSO and succinylacetone-treated Friend cells DMSO

DMSO+SA

Incubation Time (hours)

Enzyme act. (A)

Immunoreactive Prot. (B)

Ratio B/A

Immunoreactive Prot.

0 24 48 72 96

9 13 33 37 31

650 916 2 310 2 580 2 350

72 70 70 69 15

1500 2200 2800 2 650

SA-treated cells are shown in table 2, together with the enzyme activity assayed during the course of induction with DMSO alone. It appeared that there was a good correlation between enzyme activity and immunological activity with a constant ratio of 71+2. Moreover, when SA was present in the medium during induction, the immunological activity increased to the same extent as without SA, suggesting that inhibition of ALA dehydrase had no effect on its own de novo synthesis. Reversion of the Effect of SA by Hemin Control experiments were performed to test the effect of hemin (low5 M) or of a mixture of hemin and DMSO on the heme pathway enzymes. No induction in ALA synthetase activity was observed in both conditions (fig. I). However, hemin induced PBG deaminase activity and potentiated DMSO effect (fig. 2). To assess if the changes observed in the enzyme activities in the presence of SA were a consequence of intracellular heme depletion, exogenous heme was added to the medium. The potentiation by SA of the DMSO induction of ALA synthetase was completely repressed 24 h after addition of hemin to the medium (fig. 1). Furthermore, the presence of hemin throughout the experiment with SA completely prevented DMSO-mediated induction of ALA synthetase activity. In contrast, the non-inducibility of PBG deaminase in the presence of SA could not be overcome by addition of exogenous hemin (fig. 2). Effect of SA and SA-Pyrrole on PBG Deaminase Activity In Vitro Brumm et al. [25] reported that in vivo and in vitro SA can condense nonenzymatically with ALA to form an SA-pyrrole resembling PBG. We decided to determine if this pyrrole ring could inhibit PBG deaminase activity, thus hindering the detection of any increase in the enzyme activity: SA or SA-pyrrole had almost no effect on the PBG deaminase activity of a normal red blood cell hemolysate (table 3). After 10 min of incubation with the SA-pyrrole (at concentrations ranging between 1.5 and 150 PM), 85 % of the initial enzyme activity was EXP Cell Res 154 (1984)

Succinylacetone and heme synthesis in Friend cells Table 3. Effects of succinylacetone and succinylacetone-pyrrole aminase activity li-eatment None SA SA-Pyrrole

Cont. (pm)

Activity (pmoles URO/h/mg Hb)

150 150 1.5 1.5

182 130 156 152 138

481

on PBG de-

A normal red blood cell hemolysate was incubated with SA or SA-pyrrole at various final concentrations for 10 min at room temperature. The remaining enzyme activity was assayed as described under Methods.

recovered. These results indicate that non-inducibility of the PBG deaminase activity by DMSO in the presence of SA cannot be accounted for by inhibition by SA or SA-pyrrole. Quantitation of Zmmunoreactive PBG Deaminase The finding that PBG deaminase activity was not inhibited by SA-pyrrole did not exclude the possibility that some catalytically inactive protein could accumulate during the induction by DMSO in the presence of SA. Therefore, an immunotitration of the enzyme was carried out, using the ability of anti-human PBG deaminase antibodies to cross-react with mouse enzyme. In the conditions described above, immunoglobulins in a fourfold dilution resulted in a 50 % inactivation of the initial PBG deaminase activity of a Friend cell lysate. Immunotitration was performed on lysates from DMSO or DMSO- and SA-treated cells adjusted to the same PBG deaminase activity (fig. 4). The identity of immunotitration curves implies that no immunoreactive non-catalytic protein is present in SAtreated cells. These results suggest that inhibition of ALA dehydrase by SA prevents the increase in PBG deaminase synthesis which is normally observed when cells are treated by DMSO alone (see fig. 2). DISCUSSION Friend erythroleukemia cells exhibited increased activities of ALA synthetase, ALA dehydrase, PBG deaminase and ferrochelatase, when treated with 1.5% DMSO, as has been described previously [ll]. However, it may be noticed that the time course of increase of ferrochelatase activity does not significantly differ from that of the early enzymes of the pathway. As soon as 24 h after DMSO treatment ferrochelatase activity was reproducibly increased over the level of untreated cells. These data imply that the delayed increase of 59Fe incorporation into heme, which has been reported to occur after DMSO treatment, is not related to late induction of ferrochelatase activity as previously suggested [ 1I]. Exb Cell Res 154 (1984)

482 Beaumont

et al.

Succinylacetone is a potent inhibitor of ALA dehydrase. In agreement with previous reports [ 181we found that maximal inhibition of the enzyme activity was obtained in Friend cells when SA was added to the culture medium at a concentration of 0.5 mM. Under these conditions, DMSO-induced hemoglobin accumulation is completely prevented and this effect is partially reversed by exogenous hemin (table 1). This suggests that SA efficiently blocks heme synthesis thus preventing hemoglobin formation. SA was further used to investigate the effects of heme synthesis inhibition on DMSO-induced expression of the heme pathway enzymes. A pronounced effect of SA was observed on the DMSO-mediated induction of ALA synthetase activity leading to a sixfold increase over a normally-induced level after 4 days of culture. This finding suggests that heme exerts a negative control over the induction of ALA synthetase by DMSO. This possibility is further substantiated by the fact that exogenous hemin reverses the effect of SA on the induction of ALA synthetase activity (fig. 1). This is the first demonstration of the existence of a negative control by heme in the differentiating erythroid system. In liver cells, there is now firm evidence that ALA synthetase is under strict feedback regulation by heme at both transcriptional [30, 311 and translational steps [32]. In addition, heme also inhibits the transfer of ALA synthetase from the cytosol into the mitochondrion [33, 341. We cannot infer from our results which mechanism is implicated in the negative control exerted by heme on the induction of ALA synthetase in the erythroid cells. However, it is tempting to speculate that it is the heme which has been shown to accumulate in the nuclei after 2 or 3 days in the presence of DMSO [35], which limits the activation of the ALA synthetase gene. To determine the effect of heme depletion on the induction of ALA dehydrase, a specific enzyme immunoassay was used to determine the amount of immunoreactive ALA dehydrase present in the cells. With DMSO alone, a constant ratio of enzyme activity to immunoreactive protein was obtained throughout the induction experiment, thus validating the results obtained with this enzyme immunoassay. In the presence of DMSO and SA, ALA dehydrase exhibited a time-course of induction similar to that of control cells treated with DMSO alone (table 2), indicating an apparent absence of control over ALA dehydrase synthesis either by heme or by intermediate metabolite of the pathway. Surprisingly, DMSO-mediated increase of PBG deaminase activity was completely abolished by SA and the enzyme level remained in the range of uninduced cells (fig. 2). This effect could not be attributed to inhibition of PBG deaminase by SA or its derivative SA-pyrrole since (i) in vitro experiments showed that these products had no direct effect on PBG deaminase activity (table 3). (ii) Immunotitration of PBG deaminase using specific antibodies indicated the absence of any cross-reacting material in DMSO-and SA-treated cells; the immunotitration curve was superimposable to that obtained with cells treated with DMSO alone (fig. 4). Since hemin by itself has been shown to induce PBG deaminase activity in Erp Cell Res I54 (1984)

Succinylacetone and heme synthesis in Friend cells

483

Friend cells, we considered the possibility that SA might prevent PBG deaminase induction with DMSO by depleting intracellular heme. However, if supplementation of the culture medium with hemin was efficient in potentiating DMSO induction of PBG deaminase, we found that hemin was unable to reverse the effect of SA on this enzyme (fig. 2). These findings seemed to be in keeping with the ‘cascade hypothesis’ first proposed by Sassa [ll] who raised the possibility that in the course of the treatment of Friend cells with DMSO, the product of one enzyme of the heme biosynthetic pathway could act as an inducer of the next enzyme. Therefore, we tested if the absence of intermediates later than ALA in SA-treated cells prevented DMSO-mediated induction of some enzymes located after the metabolic block. Direct measurement of ferrochelatase activity (fig. 3) revealed that the enzyme had the same pattern of induction in the absence or presence of SA, thus arguing against the ‘cascade hypothesis’. However, the non-inducibility of PBG deaminase when PBG formation is impaired raises the hypothesis that PBG plays a necessary role in PBG deaminase gene expression. Alternatively, PBG may prevent PBG deaminase degradation, since it has been shown to form covalent complexes with this enzyme [27]. Further experiments including quantitation of PBG deaminase mRNA are currently in progress in our laboratory to clarify this point. The authors thank Dr P. E. Tambourin (Institut Curie, Orsay, France) for kindly providing Friend cells (clone 745) and Mrs C. Guyomard for typing the manuscript. This work was supported in part by research grant CRL 82.7007 from the Institut National de la Sante et de la Recherche Medicale and grants from the University of Paris VII.

REFERENCES 1. Elder, G H, Iron in biochemistry and medicine (ed A Jacobs & M Worwood) vol. 2, p. 243. Academic Press, London and New York (1980). 2. Kappas, A, Sassa, S & Anderson, K E, The metabolic basis of inherited disease (ed J B Stanbury, J B Wyngaarden, D S Fredrickson, J L Goldstein & M S Brown) 5th edn, p. 1301. Academic Press, New York (1983). 3. Bruns, G P & London, I M, Biochem biophys res commun 18 (1965) 236. 4. Ponka, P & Neuwirt, J, Blood 33 (1969) 690. 5. Karibian, D & London, I M, Biochem biophys res commun 18 (1965) 243. 6. Ibrahim, N G, Gruenspecht, N R & Freedman, M L, Biochem biophys res commun 80 (1978) 722. 7. Bonanou-Tzedaki, S A, Sohi, M & Amstein, H R V, Cell diff 10 (1981) 267. 8. Monette, F C & Holden, S A, Blood 60 (1982) 527. 9. Lu, L & Broxmeyer, H E, Exp hematol 11 (1983) 721. 10. Friend, C, Scher, W, Holland, J G & Sato, T, Proc natl acad sci US 68 (1971) 378. 11. Sassa, S, J exp med 143 (1976) 305. 12. Eisen, H, Bach, R & Emery, R, Proc natl acad sci US 74 (1977) 3898. 13. Chen, Z X, Banks, J, Ritkind, R A & Marks, PA, Proc natl acad sci US 79 (1982) 471. 14. Shen, D W, Real, FX, Delco, A B, Old, L J, Marks, PA & Rifkind, R A, Proc natl acad sci US 80 (1983) 5919. 15. Lowenhaupt, K & Lingrel, J B, Proc natl acad sci US 76 (1979) 5173. 16. Gusella, G F, We& S A, Tsiftsoglou, A S, Volloch, V, Neumann, J R, Keys, C & Housman, D E, Blood 56 (1980) 481. Exp Cell Res 154 (1984)

484 Beaumont et al. 17. Granick, J L & Sassa, S, J biol them 253 (1978) 5402. 18. Ebert, P S, Hess, R A, Frykholm, B C & Tschudy, D P, Biochem biophys res commun 88 (1979) 1382. 19. Schoenfeld, N, Greenblat, Y, Epstein, 0 & Atsmon, A, Biochim biophys acta 721 (1982) 408. 20. Giger, V & Meyer, U A, FEBS lett 153 (1983) 335. 21. de Matteis, F & Marks, G S, FEBS lett 159 (1983) 127. 22. Strand, L J, Swanson, A L, Manning, J, Branch, S & Marver, H S, Anal biochem 47 (1972) 457. 23. Deybach, J C, de Verneuil, H & Nordmann, Y, Human genet 58 (1981) 425. 24. Conscience, J F, Miller, R A, Henri, J & Ruddle, F H, Exp cell res 105 (1977) 401. 25. Brumm, P J & Friedmann, H C, Biochem biophys res commun 102 (1981) 854. 26. Anderson, P M & Desnick, R J, J biol them 254 (1979) 6924. 27. - Ibid 255 (1980) 1993. 28. Mage, M G, Methods in enzymology (ed H Van Vunakis & J L Langone) vol. 70, p. 142. London (1980). 29. Avrameas, S & Temynck, T, Immunochemistry 8 (1971) 1175. 30. Srivastava, G, Brooker, J D, May, B K & Elliott, W H, Biochem j 188 (1980) 781. 31. Yamamoto, M, Hayashi, N & Kikuchi, G, Biochem biophys res commun 105 (1982) 985. 32. - Ibid 115 (1983) 225. 33. Ades, I, Biochem biophys res commun 110 (1983) 42. 34. Hayashi, N, Watanabe, N & Kikuchi, G, Biochem biophys res commun 115 (1983) 700. 35. Lo, S, Aft, R & Mueller, G C, Cancer res 41 (1981) 864. Received February 2 1, 1984

Exp Cell Res 154 (1984)

Printed

in Sweden